I don't think anyone doubts that when we solve problems on platforms like Codeforces, we are primarily sharpening our mathematical thinking, and only secondarily our algorithmic skills (such as implementation, data structures, etc.). However, lately, I’ve noticed a rather controversial trend: sometimes problemsetters don’t seem to put much effort in and just cram pure mathematics into a basic <input/output> format. You've probably encountered this in recent rounds: you open a problem, and its entire "solution" boils down to guessing a formula or a pattern just by looking at the samples.

By no means am I denying the benefits of math, and I love it myself. But the context in which it is currently being served leaves much to be desired. I’m absolutely not against problems where proving a specific fact serves as the foundation for the solution (for instance, optimizing a brute-force search).

Instead, I suggest adding more variety to the problems—incorporating things like "olympiad tricks" into the implementation itself. This way, the mathematical idea would only be half the battle, while the other half would depend on how you implement it. Such an approach would maintain a healthy balance and keep everyone happy: mathematics wouldn’t go anywhere, while the implementations would become far more interesting and engaging.

In the last few years, there have been significantly fewer problems involving mathematics and number theory. I don't understand why this is happening, as there always seemed to be plenty of mathematicians in the competitive programming community. Perhaps problem-setters think that problems involving numbers have trivial code, but sometimes, using number-theoretic methods, interesting and non-trivial optimizations can be achieved.